Randomized trial · people
Semaglutide (oral) did not slow early Alzheimer's in evoke trials
Checked against the paper Randomized trialPublished by PepVault
The takeaway
In these two large trials, oral semaglutide did not slow early Alzheimer's disease, and its safety was in line with what has been seen in its other uses.
- Kind
- Randomized trial
- Studied in
- People
- Number studied
- 3808
- Published
- 2026
- Journal
- Lancet
What they tested
The team tested whether daily oral semaglutide could slow the worsening of early Alzheimer's disease compared with a placebo, a dummy pill.
How the study was done
Two randomized, double-blind, placebo-controlled phase 3 trials enrolled 3808 people aged 55-85 with amyloid-confirmed Alzheimer's disease and mild cognitive impairment or mild dementia, across 566 sites in 40 countries. People were assigned 1:1 to oral semaglutide 14 mg once daily (flexible dose) or placebo for up to 156 weeks, and the main result was measured at week 104.

Model. Individual results vary.
What they found
The main measure was CDR-SB, a dementia rating score where higher means worse, and by week 104 it had risen by about the same amount in both groups. In evoke, the score went up 2.3 on semaglutide and 2.3 on placebo, a gap of -0.08 points with a p-value of 0.57 and a 95% confidence interval running from -0.35 to 0.20. In evoke+, it went up 2.2 on semaglutide and 2.1 on placebo, a gap of 0.10 points with a p-value of 0.46 and a 95% confidence interval running from -0.17 to 0.38. Side effects that came up during treatment were seen in 1729 (91.2%) of 1896 people taking semaglutide and 1613 (84.8%) of 1902 taking placebo, and investigators linked five deaths to treatment: one on semaglutide and four on placebo.
What it doesn’t show
The trials tested only the 14 mg flexible dose of the oral form, in people with early, amyloid-confirmed Alzheimer's, and only 54 (2.8%) people in evoke+ had small vessel pathology. Both trials were stopped because of the negative result, and they were funded by Novo Nordisk.
The paper
Cummings et al. Efficacy and safety of oral semaglutide 14 mg (flexible dose) in early-stage symptomatic Alzheimer's disease (evoke and evoke+): two phase 3, randomised, placebo-controlled trials. Lancet. 2026.
The compounds
More studies on Semaglutide (oral)
- Semaglutide (oral): smaller HbA1c drop than orforglipron in ACHIEVE-3Randomized trial · people · 2026
- Semaglutide 25 mg pill: 13.6% weight loss at 64 weeks in OASIS 4Randomized trial · people · 2025
- Semaglutide pill: 15.1% weight loss at 68 weeks in OASIS 1Randomized trial · people · 2023
- Semaglutide (oral): heart risk not higher than placebo in PIONEER 6Randomized trial · people · 2019
- Every study on Semaglutide (oral)
Researched and drafted with AI, then checked claim by claim against the primary sources linked below by a separate review pass. Published by PepVault. Last reviewed . Spot an error? feedback@pepvault.app. How we research. This page reports one study; it is not medical advice.